The premise of this topic is that every game designed for competitive play has a baseline level of vocabulary a person needs to possess to do a competent game call.
I believe being aware of this, and choosing vocabulary is actually a very relevant game design topic that needs to be exposed for developers to give their future projects the best chance of building a fanbase.
So lets realize this concept.
Now for Starcraft, you'd be in trouble if you didn't know the names of races, units, and abilities and tried to cast a game. Immediately you'd be amiss in trying to describe the action reasonable without these things.
"Shiny yellow armor alien man with laser blades, running towards lizard dog monsters, slashing them to bits! There are no more dog monsters! Clearly the alien men have pulled ahead in this match."
Bit of an odd way to describe a zealot picking off a few zerglings. But if you took someone with zero knowledge of the game's details that is essentially how it would go (I may or may not be exaggerating.)
Beyond this, things get more extraneous. Certainly it would be useful to know that the phoenix range upgrade is called the 'Anion Pulse-Crystals', but using this piece of lingo has diminished returns compared to just learning basic knowledge. It would add polish to a cast and make you look super cool, but the use of that term over just saying 'He is researching Phoenix Range Upgrade!' does not have the same yields.
As well in this scenario, the knowledge of this title might not even be possessed by the audience which ultimately affects its usage and flow too. Key point. If you have to add on, that Anion Pulse Crystals is the Phoenix Range Upgrade everytime, it becomes less efficient. In faster games, where everything you say may or may not be important this is an unfortunate circumstances.
Terms beyond the basic that a caster could use also extend to lingo devised by the community itself. A Starcraft example would be an 'Archon Toilet' or 'Forcefield Doughtnut'. (I won't inquire into why these terms are so ...trashy).
Ultimately though a developer probably can not take control of the development of language around their game post release, and while they can approve it by incorporation later, it is not something relevant to the creation of the game since it would occur afterwards. With the mild note, that language would usually only develop if something occurred unforseen to the developers that needed a descriptor, so perhaps they should have a bit more foresight, if they truly wanted to control language.
So that type of language is being ignored for this exercise.
Now a game like Starcraft actually is fairly simple in terms of the vernacular or dialect of the game itself. There's maybe ~200 units, abilities, upgrades, structures, and miscellaneous terms unique to the game.
RTS in the late 90s to early 2000s was notable for a lot of 'gunk'. The blizzard philosophy of streamlined unit relationships and consistent rules was part of the reason behind the success of the -Crafts.
Players could grasp Starcraft BroodWar despite its strategic complexities because they could get behind the relationship between lings and marines, whereas other RTS had gravitated down the route of tesla coil wielding jetpack gorillas receiving a +4 bonus to their banana divekicks when your windows clock ended in an odd number.
This simplicity shows up in the language around the game too. This has the beautiful affect of making the game accessible to shoutcasters. The point of entry is rather low. You don't have to do much homework, to do a simple play-by-play.
That is why through the first years of Starcraft you saw massive amounts of people try their hat at casting, to various results. (homework question: count how many attempted then escaped the scene in the last bit of time since release).
Now this isn't a single effect, the popularity, and the type of popularity Starcraft has (a stable long cycle) affects the last point greatly. So games which expect a longer consumption time can be a bit more brazen with their vocabulary choice. But I will come back to this.
So other genres nowadays show a lot of unnecessary complexity, this makes their vernacular a tangled mess and creates a higher point of entry.
Let's look at Action-RTS (MOBA).
Fantastic games, but ones which do suffer from a lack of a rogue's gallery of shoutcasters.
League of Legends caster ecosystem involves a lot of first party casters who are employed directly through Riot and a smallish quantity of those who developed through the community . Given the age of the game (1+ year beyond Starcraft 2 Beta), and its relative fanbase size (BIGBIGBIG).
DotA2, a game on the GROW whose predecessor came out a shockingly long time ago (hint: DotA was actually released for the Amiga). It's problems with developing a caster scene have been well discussed and I have seen established casters like TobiWan or Purge lament a lack of viable partners previous.
Now to this point we've ignored, the motives of why someone would be interested in the media side of competitive gaming. But let's take the Chaotic Evil variety, someone decidedly in it for a popularity grab and to sell their own personalized T-Shirts.
The above examples, are begging for more warm bodies, so why wouldn't this evil individual make an attempt? Everyone knows League of Legends has more fans than the population of China! But Evil person instead decides to study Business and be evil the typical way. (there are good businessmen too! ...sometimes)
Well there are many preclusive elements that come from design, but lets stick to our topic of discussion, vernacular.
Okay so your average Action-RTS is built on having a stable of heroes to sell via micro transactions.
This means most character rosters are at least 100 heroes. Normal models are 5 abilities. Add in heaps of items, creeps, giant golems you have to slay for treasures,
These games extend well beyond our example of Starcraft and you could expect a vernacular of 600+ terms before you can even hit our baseline.
You can add in skins, muddled lore (poorly written so there is no singleplay, Demacians versus Zszzzz.), and other factors for less crucial terminology and our figure would grow even larger. (more extraneous terms than Starcraft on average)
Now take my crude example from before with the zealot vs zerglings. The type of gameplay in an Action-RTS tends to possess very high energy short bursts of action, with a lot of varied, ability-based interactions.
So that means tried to describe something without absolute knowledge would typically sound like a very gregarious mouth fart.
So where does this complexity derive from?
Namely its a push for depth from the developer. A creator's job is to make their game of course, awesome. So drawing lore, naming skills, producing this content for the casual fan is all about spinning a very deep multifaceted experience.
But we know developing a game for competitive play might now consider the viewer experience as well. Since casting is so relevant to viewer experience this must become a point of two-sided concern for developers.
One the one hand the creation of my 'Fire Throwing Demon' character in my soon to be award winning Action RTS game, might fall a little flat if I pick the following naming schemes in to enable memorization and ease of learning.
Skill 1: Throw Fire Skill 2: Throw Double Fire ... Ultimate: Throw Shit Balls Tons of Fire!
On the other hand if I gravitate down the route of the following.
Skill 1: Consummate Daemonic Pyre Projection ....
I can expect, a lot of difficulty with casters adopting our vernacular and casting the game properly. Beyond that consider our game attempting to go global, without local translations available for every region and the difficulties with non-native speakers, expect headaches from Madrid to Shanghai in our casting communities.
So what is the answer?
Well as is the normal, its likely somewhere in the grey of these two extremes, but its definitely something quite application specific in finding a resolution. As well I'm not sure I have all the answers, but identifying this problem is the first step.
There is a very special part of the competitive community which I think falls victim to this the most, and the actual inspiration for this piece.
Now fighting games tend to have a quick turnover, and a lot of forking in the community (as in many games active at once). Players move from Virtua Fighter, to Mortal Kombat, to MvC, to crazy JPOP fighting game and back. There is a lot of game switching, and the lifetime of games can tend much smaller and are generally short consumption period titles, compared to Starcraft which we established as a longer consumption period title. (fighting games have a few exceptions, either from developers not taking an active interest in the genre or truly great titles holding up over time)
So fighting games possess vernaculars that do not just make it difficult to learn the terminology, they make it impossible. The size of character rosters, the amount of abilities (and how they defy categorization), and how quickly the turnover I mentioned, all contribute to this.
Let's look at an excerpt of Voldo's movelist from Soul Calibur V.
Caliostro Rush Katar Slap Writhing Mantis Writhing Lunatic Snake Eater Writhing Fumble Life Sucker Centipede Nightmare Rat Retreat Blind Caliostro Rush Writhing Blind Claw Writhing Blind Spring Writhing Blind Spin Blind Drop Kick Writhing Blind Fumble Rat Chaser Superwyrm Superior Arachnid Superwyrm Escape Gravestone Cleaver Wheel of Madness Blade Arch Diablo Brothers Fanatic Arachnid
This is around HALF of his(?) named moves. I had to cut it there because it was just too much to list.
Fighting game commentary if you haven't had the pleasure is a different animal. A laid back, flavouring compared to Starcraft's dominant lead.
The evolution of this style of game commentary has a basis arcade culture and the personalities involved in the culture's growth. But I contend that the inability for players and commentators to fully grasp the games internal language is actually a dominant factor in its current form. If you can't describe the action literally it makes total sense why things have gone the way they have.
In typical fighting game culture fashion you might argue, hey, let it be. The spirit behind it is too interesting to mess with. But to me style, and delivery are different. Let people make choices and not be forced into them.
Let's look at cause one last time.
Recall the days of fireballs and dragon punches. Developers saw the allure in creating identities to each fighters abilities like this, so as the genre's complexity grew so too did the word lists.
Before competitive gaming took on the state it did now (streaming and livecasting), there was no reason to stop. Developers had no reason to stop chucking words, character stories, and crazy names at the wall. If one of those stuck if was all gravy. That meant more cultural impact, and more relevance for their game.
But now, this practice can actually be damaging. If you can foster a competitive community that creates dynamic content with good appeal your game intended for that space could fail.
So it's time to refine, simplify, and consider the audience.
I actually feel that Dota does the naming quite well. There are very few skills longer than 2 words and most of the skill names are related to what they do. This still means the casters ignore the skill names, not because they don't know them, but because the viewer doesn't know it.
Calling it Sven's stun is actually better than calling it Storm Bolt. Both are two words, the first one helps the new viewers, the second one doesn't add anything. There are exceptions to this when calling Void, Night Stalkers Slow (or nuke) takes much longer and removed information.
Then you have iconic skills like Blackhole and Ravage, these are used to signify their importance. Of course these things vary with casters, yet I feel a good balance as you mentioned above with the phoenix range upgrade helps in ARTS games as well.